Retinal ganglion cells (RGCs) are the primary target of glaucoma disease pathogenesis and the first cell type to develop from mitotic progenitors in the optic cup. The bHLH transcription factor Math5 is transiently expressed following cell cycle exit, and required for RGC genesis. Using Cre-lox lineage analysis and BAC transgenic mice, we have determined that 5% of adult retinal neurons derive from post-mitotic Math5+ cells, but only 1 in 9 Math5+ precursors develops into an RGC. These cells are fully multipotent and express equivalent levels of Math5 regardless of their fate choice. In Math5 mutants, the lineage-marked cells form all major cell types except RGCs. Math5 is thus a competence factor - necessary but not sufficient - for RGC fate specification. Additional positive or negative factors are needed for ganglion cell development. Notch signaling is known to inhibit RGC fate, but it is unclear whether Notch acts in progenitors or post-mitotic cells (before or after Math5), and whether Notch affects secondary fate choice in the Math5 lineage. When we activate Notch in Math5+ cells, they initially develop as retinal neurons but transdifferentiate into Muller glia. When we express Math5 in committed photoreceptors, multipotency is restored. This proposal studies the plasticity of Math5+ cells and downstream factors required to form RGCs in mice. We aim: (1) to test the hypotheses that first-born cells in the mouse retina express Math5 and develop into RGCs as a default fate, and that Notch is the major inhibitor of RGC fate in Math5+ cells, using BAC transgenes - to trace fates of Math5 descendants that receive Notch signals, and to drive or block Notch in Math5+ cells with rapid or slow kinetics, in mutant and wild-type mice - and retinal explants; and (2) to evaluate Math5 and the Nrl as permissive and instructive factors for RGC and rod fates, by cross-expressing them in photoreceptor and ganglion cell precursors.